On SRX5400/5600/5800, MX240/480/960 series platforms one SPU may stop forwarding traffic when a specific bit is not cleared by SCHEDF (fabric scheduler). In this issue since bit is not cleared, which would indicate resources available to send a packet and this is causing traffic stop on all CPU cores of the affected SPU. 

On SRX5k series with SPC3 card, in rare cases an SPU may stop forwarding traffic. In this case, logs will show a "potential pkt drop issue" on all the cores of that SPU and an FPC Major Alarm would be raised due to the PIC_CMERROR_TALUS_PKT_LOSS error being reported and:

[Mon Jun 7 21:52:08.493 LOG: Err] Core 2 has potential pkt drop issue. Has dropped 100 pkt within 900 sec

[Mon Jun 7 21:52:08.493 LOG: Err] Core 3 has potential pkt drop issue. Has dropped 100 pkt within 900 sec

[Mon Jun 7 21:52:08.493 LOG: Err] Core 4 has potential pkt drop issue. Has dropped 100 pkt within 900 sec

[Mon Jun 7 21:52:08.493 LOG: Err] Core 5 has potential pkt drop issue. Has dropped 100 pkt within 900 sec

[Mon Jun 7 21:52:08.493 LOG: Err] Core 6 has potential pkt drop issue. Has dropped 100 pkt within 900 sec

https://prsearch.juniper.net/problemreport/PR1600216https://prsearch.juniper.net/problemreport/PR1600216

RESOLVED IN

19.4R2-S6

19.4R3-S7

20.2R3-S3

20.4R3-S2

21.1R3

21.2R2

21.2R2-S1

21.2R3

21.3R2

21.4R1

Workaround is to reseat FPC